tetano
Editor, Senior Moderator
Sci Rep
. 2025 Mar 19;15(1):9459.
doi: 10.1038/s41598-025-85733-5. Predicting coronavirus disease 2019 severity using explainable artificial intelligence techniques
Takuya Ozawa[SUP] #[/SUP][SUP] 1 [/SUP], Shotaro Chubachi[SUP] #[/SUP][SUP] 2 [/SUP], Ho Namkoong[SUP] #[/SUP][SUP] 3 [/SUP], Shota Nemoto[SUP] 4 [/SUP], Ryo Ikegami[SUP] 4 [/SUP], Takanori Asakura[SUP] 1 5 6 [/SUP], Hiromu Tanaka[SUP] 1 [/SUP], Ho Lee[SUP] 1 [/SUP], Takahiro Fukushima[SUP] 1 [/SUP], Shuhei Azekawa[SUP] 1 [/SUP], Shiro Otake[SUP] 1 [/SUP], Kensuke Nakagawara[SUP] 1 [/SUP], Mayuko Watase[SUP] 1 [/SUP], Katsunori Masaki[SUP] 1 [/SUP], Hirofumi Kamata[SUP] 1 [/SUP], Norihiro Harada[SUP] 7 [/SUP], Tetsuya Ueda[SUP] 8 [/SUP], Soichiro Ueda[SUP] 9 [/SUP], Takashi Ishiguro[SUP] 10 [/SUP], Ken Arimura[SUP] 11 [/SUP], Fukuki Saito[SUP] 12 [/SUP], Takashi Yoshiyama[SUP] 13 [/SUP], Yasushi Nakano[SUP] 14 [/SUP], Yoshikazu Muto[SUP] 15 [/SUP], Yusuke Suzuki[SUP] 5 6 [/SUP], Ryuya Edahiro[SUP] 16 [/SUP], Koji Murakami[SUP] 17 [/SUP], Yasunori Sato[SUP] 18 [/SUP], Yukinori Okada[SUP] 16 19 20 [/SUP], Ryuji Koike[SUP] 21 [/SUP], Makoto Ishii[SUP] 1 22 [/SUP], Naoki Hasegawa[SUP] 23 [/SUP], Yuko Kitagawa[SUP] 24 [/SUP], Katsushi Tokunaga[SUP] 25 [/SUP], Akinori Kimura[SUP] 26 [/SUP], Satoru Miyano[SUP] 27 [/SUP], Seishi Ogawa[SUP] 28 29 [/SUP], Takanori Kanai[SUP] 30 [/SUP], Koichi Fukunaga[SUP] 1 [/SUP], Seiya Imoto[SUP] #[/SUP][SUP] 31 [/SUP]
Affiliations
Predictive models for determining coronavirus disease 2019 (COVID-19) severity have been established; however, the complexity of the interactions among factors limits the use of conventional statistical methods. This study aimed to establish a simple and accurate predictive model for COVID-19 severity using an explainable machine learning approach. A total of 3,301 patients ≥ 18 years diagnosed with COVID-19 between February 2020 and October 2022 were included. The discovery cohort comprised patients whose disease onset fell before October 1, 2020 (N = 1,023), and the validation cohort comprised the remaining patients (N = 2,278). Pointwise linear and logistic regression models were used to extract 41 features. Reinforcement learning was used to generate a simple model with high predictive accuracy. The primary evaluation was the area under the receiver operating characteristic curve (AUC). The predictive model achieved an AUC of ≥ 0.905 using four features: serum albumin levels, lactate dehydrogenase levels, age, and neutrophil count. The highest AUC value was 0.906 (sensitivity, 0.842; specificity, 0.811) in the discovery cohort and 0.861 (sensitivity, 0.804; specificity, 0.675) in the validation cohort. Simple and well-structured predictive models were established, which may aid in patient management and the selection of therapeutic interventions.
Keywords: Artificial intelligence; COVID-19; Machine learning.
. 2025 Mar 19;15(1):9459.
doi: 10.1038/s41598-025-85733-5. Predicting coronavirus disease 2019 severity using explainable artificial intelligence techniques
Takuya Ozawa[SUP] #[/SUP][SUP] 1 [/SUP], Shotaro Chubachi[SUP] #[/SUP][SUP] 2 [/SUP], Ho Namkoong[SUP] #[/SUP][SUP] 3 [/SUP], Shota Nemoto[SUP] 4 [/SUP], Ryo Ikegami[SUP] 4 [/SUP], Takanori Asakura[SUP] 1 5 6 [/SUP], Hiromu Tanaka[SUP] 1 [/SUP], Ho Lee[SUP] 1 [/SUP], Takahiro Fukushima[SUP] 1 [/SUP], Shuhei Azekawa[SUP] 1 [/SUP], Shiro Otake[SUP] 1 [/SUP], Kensuke Nakagawara[SUP] 1 [/SUP], Mayuko Watase[SUP] 1 [/SUP], Katsunori Masaki[SUP] 1 [/SUP], Hirofumi Kamata[SUP] 1 [/SUP], Norihiro Harada[SUP] 7 [/SUP], Tetsuya Ueda[SUP] 8 [/SUP], Soichiro Ueda[SUP] 9 [/SUP], Takashi Ishiguro[SUP] 10 [/SUP], Ken Arimura[SUP] 11 [/SUP], Fukuki Saito[SUP] 12 [/SUP], Takashi Yoshiyama[SUP] 13 [/SUP], Yasushi Nakano[SUP] 14 [/SUP], Yoshikazu Muto[SUP] 15 [/SUP], Yusuke Suzuki[SUP] 5 6 [/SUP], Ryuya Edahiro[SUP] 16 [/SUP], Koji Murakami[SUP] 17 [/SUP], Yasunori Sato[SUP] 18 [/SUP], Yukinori Okada[SUP] 16 19 20 [/SUP], Ryuji Koike[SUP] 21 [/SUP], Makoto Ishii[SUP] 1 22 [/SUP], Naoki Hasegawa[SUP] 23 [/SUP], Yuko Kitagawa[SUP] 24 [/SUP], Katsushi Tokunaga[SUP] 25 [/SUP], Akinori Kimura[SUP] 26 [/SUP], Satoru Miyano[SUP] 27 [/SUP], Seishi Ogawa[SUP] 28 29 [/SUP], Takanori Kanai[SUP] 30 [/SUP], Koichi Fukunaga[SUP] 1 [/SUP], Seiya Imoto[SUP] #[/SUP][SUP] 31 [/SUP]
Affiliations
- PMID: 40108236
- DOI: 10.1038/s41598-025-85733-5
Predictive models for determining coronavirus disease 2019 (COVID-19) severity have been established; however, the complexity of the interactions among factors limits the use of conventional statistical methods. This study aimed to establish a simple and accurate predictive model for COVID-19 severity using an explainable machine learning approach. A total of 3,301 patients ≥ 18 years diagnosed with COVID-19 between February 2020 and October 2022 were included. The discovery cohort comprised patients whose disease onset fell before October 1, 2020 (N = 1,023), and the validation cohort comprised the remaining patients (N = 2,278). Pointwise linear and logistic regression models were used to extract 41 features. Reinforcement learning was used to generate a simple model with high predictive accuracy. The primary evaluation was the area under the receiver operating characteristic curve (AUC). The predictive model achieved an AUC of ≥ 0.905 using four features: serum albumin levels, lactate dehydrogenase levels, age, and neutrophil count. The highest AUC value was 0.906 (sensitivity, 0.842; specificity, 0.811) in the discovery cohort and 0.861 (sensitivity, 0.804; specificity, 0.675) in the validation cohort. Simple and well-structured predictive models were established, which may aid in patient management and the selection of therapeutic interventions.
Keywords: Artificial intelligence; COVID-19; Machine learning.